Questions 11

  1. What are the consequences of omitting a significant variable in the model?
  2. In what special case we can obtain valid estimates of parameters even so significant variables were omitted from the model?
  3. Why the variables which are not significant should be deleted from the model?
  4. Parameters related to variables {$x_1$} and {$x_2$} are positive. The variables are negatively correlated. What will be the effect of omitting variable {$x_1$} in the regression on the estimate of the parameter related to variable {$x_2$}?'
  5. Explain when we are saying that the observation is unusual and when we can identify an observation as erroneous.
  6. In what case an unusual observation has a large impact on regression results?
  7. What statistics are use to detect unusual and erroneous observations?
  8. In what case we are talking about perfect collinearity? How we are solving this problem?
  9. What are the consequances of imperfect collinearity? What statistics is used to detect the presence of severe collinearity in the model?